Is there a way to distinguish nouns and verbs that applies to all languages?

This problem has been occupying my mind for some time now. I'm not quite sure how to approach this question, so I'll just list out some of my thoughts that make this question difficult:

  • The usual grade school definition of a noun is a person, place, or thing, while a verb is defined as an action or state of being. This doesn't work so well when confronted with words like reenactment which is a noun that clearly refers to an action, or to locate which seems like it refers to a concept of place.
  • Some words can clearly fall in both categories, like struggle, which is both noun and verb at once. (Though my belief is that the categories can exist independently, even if many words cross into both categories.)
  • What's the difference between a language that has verbs like "to be a book", and languages that allow you to string nouns together making "X Y" the same meaning as "X is Y"?

The general feel I'm getting from all this is that nouns and verbs are distinguished by their roles in a sentence, not from the concepts they refer to, but I'm having a hard time formalizing this.

Follow-up: It seems like people are getting thrown off by all the thoughts I listed out above. Here's another way of restating the question:

If you have two classes of words in a language, how do you know which one is nouns and which one is verbs?

  • 1
    How would this apply to polysynthetic languages I wonder. Aren't there some where a single individual word can be made up of part noun and part verb? – hippietrail Nov 20 '11 at 13:38
  • @ Joe, When you've already established two types of word classes (based on formal criteria), then you need to apply semantic and pragmatic criteria (see my answer below). Remember, one criterion or two or even three criteria are never enough; you need to apply all four of them. – Alex B. Nov 22 '11 at 17:18

Where do I start? At first, there are literally tons of research on this topic (some call them parts of speech, others call them word classes). I don't know how much time you're willing to spend on reading or how linguistically well-trained you are.

To be on the safe side, I strongly recommend to start with an excellent review article by Walter Bisang, Word Classes, In The Oxford Handbook of Linguistic Typology, ed. J. J. Sung, pp. 280-302. Oxford: OUP, 2010.

Main points summarized here:

  1. The inadequacy of purely semantic/notional definitions.
  2. Four prerequisites for distinguishing word classes (you need all four of them):

- semantic criteria:

Sasse 1993: nouns are thing-like concepts, verbs are event-like concepts;

Langacker 1987: nouns are static and holistic, verbs are dynamic;

Givon 1979: nouns represent ontological categories that are stable in time, unlike verbs (very time-unstable).

Wierzbicka 2000, Dixon 2004: there are certain semantic types that are always associated with nouns or verbs only, e.g. PEOPLE, PARTS, FLORA etc. are always nouns, whereas MOTION, SPEAKING are always verbs.

Croft 2000: nouns refer to objects, verbs express predication of an action etc.

- pragmatic criteria/criteria of discourse functions:

Nouns introduce participants, verbs assert the occurrence of an event.

- formal criteria:

Nouns and verbs have different morphological and syntactic distribution. They may also differ in their phonological form.

- distinction between lexical and syntactic levels of analysis:

lexical (paradigmatic) vs. syntactic (syntagmatic) levels.

He also talks about approaches to word classes proposed by Schachter 1985, Hengeveld 1992, Croft (in a series of works). There is a special section devoted to the noun/verb distinction (section 5). There is a nice conclusion there, where he cites three criteria (from Evans and Osada 2005) which a language lacking a noun/verb distinction must meet:

  • compositionality;
  • bidirectionality;
  • exhaustiveness.
| improve this answer | |
  • Mothion is noun. – Anixx Nov 24 '13 at 17:17
  • 3
    @Anixx, MOTION (all caps) is not a noun, it's a concept that can be realized as "to go," "to run," "to swim" etc. – Alex B. Nov 24 '13 at 17:20
  • What does "holistic" and "dynamic" and "bidirectional" and "exhaustive" even mean? – Pacerier Feb 16 '18 at 0:53
  • @AlexB., Are you sure "The Oxford Handbook of Linguistic Typology" is not focused on English language? – Pacerier Feb 16 '18 at 0:56
  • 1
    @Pacerier I'm pretty sure it's not because I've read it. Feel free to peruse it yourself. – Alex B. Feb 16 '18 at 3:46

The modern syntax approach to this, unlike the traditional grammar approach, is to categorize words according to morphological and syntactic properties.

Morphologically, we can observe that nouns can be derived from other nouns using the suffix -ism, for example, to indicate a belief or process (capital > capitalism). Verbs are derived from other kinds of affixes, such as -ify (terror > terrify) and un- (do > undo). Besides, nouns can be inflected for number (book > books) and verbs can be inflected for tense, aspect and mood (give > given).

Syntactically, a noun can be placed after a definite article (technically, a determiner), for example (the book is green), but not after an adverb like more (*this one is more book than the other). A verb can be followed by an adverb of manner, like humbly (he speaks humbly), but not by an article (*a speaks is fine). Likewise, if you can substitute a word with a regular noun or verb, both words will belong to the same category.

The combination of morphological and syntactic analyses allows you to define what category a word belongs to. About the third item, you are referring to a specific use of the verb to be, known as copula, more specifically, the zero copula, present in languages such as Russian and Hebrew.

| improve this answer | |
  • 4
    If I'm understanding your answer correctly (and it's quite possible that I'm not), it merely pushes the problem of distinguishing verbs from nouns onto other morphemes. The above morphological point seems to only show that there are two classes of words in a particular language, not show which one is nouns and which is verbs. The syntactic point presupposes that we have a good definition of "article" and "adverb" that don't rely on already distinguishing "noun" from "verb". – Joe Oct 10 '11 at 19:39
  • 4
    The problem I face is defining "noun" and "verb" universally, not just for any particular language. I don't see any way to extend this "-ism" type of definition to all languages (though it's spot-on for English). And I don't think it's universal that only verbs inflect for tense/aspect and only nouns inflect for number. – Joe Oct 10 '11 at 21:13
  • 5
    @Joe Otavio Macedo's answer is good, but I have to disagree with his comment, the morphosyntactic evidence is entirely reliable and is in fact the primary way of determining the lexical categories within a language. Having done that we can then choose an appropriate name for a category (verb/noun) by looking at the members of the categories. That category whose members are used to represent actions and states, and which contains words equivalent to English 'hit/eat/etc' can be called 'verbs', while the category that refers to time-stable entities (rocks/animals/etc) can be called 'nouns'. – Gaston Ümlaut Oct 10 '11 at 23:28
  • 3
    @Joe Re universal definition of verb/noun. It's problematic trying to define a lexical category by semantics. It does seem that every language has at least the two categories, which can be named 'noun' and 'verb' because they each contain the appropriate prototypical elements, but this can't be used to define the categories, only to characterise them. – Gaston Ümlaut Oct 10 '11 at 23:37
  • 4
    This position is very Anglocentric. There are languages without articles, there are languages where nouns can be marked for tense; there are languages with a very flexible word order etc. – Alex B. Nov 17 '11 at 17:40

If you'd like to research this question more in-depthly, you should check out Mark Baker's (2002) book Lexical Categories. There he focuses on languages for whom the difference between nouns and verbs (and adjectives) are less straightforward. But it requires a fairly advanced background in syntactic theory.

| improve this answer | |
  • A quick summary of essential points since you've read it? – Pacerier Feb 16 '18 at 1:02

The references in @Alex B.'s answer will do the job for a summary. Van Valin 2008 and ch.2 of Croft 2001 "Radical Construction Grammar" might also be useful. I'll try to summarize a family of views held by typologically oriented-linguists, but I couldn't guarantee that anyone other than myself accepts all of the statements made here.

  1. It is possible to find categories corresponding to nouns and verbs in different languages, but the approach will be a sophisticated version of the traditional grammar view. The prototypical noun is a referring expression (whence "person, place, thing"), and the prototypical verb is a predicating expression (whence "action"). It is then usually possible to pick out morphosyntactic properties associated with referring constructions and predicating constructions.
  2. But there are many languages where predicating and referring constructions do not always clearly distinguish two classes of words (i.e. "nouns" can also predicate, "verbs" can also refer). (Van Valin 2008 gives examples of this type of behavior from Nootka and Tagalog) However, it will be possible to find other types of constructions which can assist with the sorting. For example, I might find a language where an intransitive predicating construction does not make the distinction between nouns and verbs, but a transitive predicating construction does. Therefore, if you know roughly which types of word classes you are looking for, it is usually possible to pick out the right constructions that will give you the formal basis for distinguishing the desired number of word classes. Since the procedure is somewhat subjective, linguists will end up disagreeing on, say, whether a language has adjectives, since those who want adjectives can keep looking until they find the construction that distinguishes adjectives from intransitive verbs.

The answer to the question is that "noun" and "verb" are comparative categories which are essential for typological studies, and all languages provide some formal basis for distinguishing them; The catch is that you have to start from meaning-based comparative categories, and the formal basis for distinguishing categories is subjective and language-specific. The chapter in the WALS on the order of Noun and Adjectives (Dryer 2011), for example, includes a large number of languages for which most constructions do not distinguish adjectives and intransitive verbs; instead, "adjective" is understood as "modifying expression". An approximate meaning-based concept of "adjective" is needed if comparisons are to be made across languages of different grammatical profiles.

| improve this answer | |

The Salishan language family (northwestern USA, southwestern Canada) is a famous locus of the debate about noun versus verb and lexical categories in general. The neighbouring Wakashan languages on Vancouver Island pose similar problems. Here are a few relatively recent references chronicling the arguments.

  • Kinkade, M. Dale. 1983. Salish evidence against the universality of ‘noun’ and ‘verb’. Lingua 60.1: 25–40.
  • van Eijk, Jan P. & Hess, Thom. 1986. Noun and verb in Salish. Lingua 69: 319–331.
  • Renker, Ann M. 1987. Rethinking noun and verb: An investigation of AUX in a Southern Wakashan language. Washington DC: American University, PhD dissertation.
  • Birch, Barbara M. 1993. Another look at Salish nouns and verbs. Papers for ICSNL XXVIII: 19–26.
  • Jelinek, Eloise & Demers, Richard A. 1994. Predicates and pronominal arguments in Straits Salish. Language 70: 697–736.
  • Demirdache, Hamida & Matthewson, Lisa. 1995. On the universality of syntactic categories. Proceedings of NELS 25: 79–94.
  • Mattina, Nancy. 1996. Aspect and category in Okanagan word formation. Burnaby BC: Simon Fraser University, PhD dissertation.
  • Burton, Strang. 1997. Past tense on nouns as death, destruction, and loss. Proceedings of NELS 27: 65–77.
  • Haag, Marcia. 1998. Word-level evidence for lexical categories in Salishan languages. International Journal of American Linguistics 64: 379–393.
  • Davis, Henry & Matthewson, Lisa. 1999. On the functional determination of lexical categories. Revue Québecoise de linguistique 27.2: 29–69.
  • Davidson, Matthew. 2002. Studies in Southern Wakashan (Nootkan) grammar. Buffalo, NY: SUNY Buffalo, PhD dissertation.
  • Davis, Henry. 2002. Categorial restrictions in Stʼátʼimcets (Lillooet) relative clauses. Papers for ICSNL XXXVII: 61–76.
  • Nordlinger, Rachel & Sadler, L. 2002. Nominal tense in cross-linguistic perspective. Language 80: 776–806.
  • Montler, Timothy. 2003. Auxiliaries and other categories in Straits Salishan. International Journal of American Linguistics 69.2: 102–134.
  • Stonham, John. 2004. Linguistic theory and complex words: Nuuchahnulth word formation, pp. 54–63. London: Palgrave Macmillan.
  • Waldie, Ryan. 2004. Nuu-chah-nulth denominal verbs. Victoria BC: University of Victoria, master’s thesis.
  • Davis, Henry. 2005. Constituency and coordination in Stʼátʼimcets (Lillooet Salish). Pp. 31–64 in Andrew Carnie, Sheila Anne Dooley, & Heidi Harley (eds.), Verb first: On the syntax of verb initial languages. Amsterdam: John Benjamins.
  • Wiltschko, Martina. 2005. The syntax of precategorial roots. Proceedings of WSCLA 2005.
  • Davis, Henry & Matthewson, Lisa. 2009. Issues in Salish syntax and semantics. Languages & linguistics compass 3.4: 1097–1166.
  • Koch, Karsten & Matthewson, Lisa. 2009. The lexical category debate in Salish and its relevance for Tagalog. Theoretical linguistics 35.1: 125–137.
  • Wiltschko, Martina. 2009. Root incorporation: Evidence from lexical suffixes in Halkomelem. Lingua 119.2: 199–233.

I think there is a newer paper by Henry Davis regarding lexical categories in Salishan but I don’t have the reference at hand.

| improve this answer | |
  • 3
    This doesn't answer the question. Rather than giving us a bibliography can you summarise what this language family brings to the discussion? – curiousdannii Nov 11 '14 at 23:47

It seems to me that the fundamental difference between nouns and verbs is that verbs form sentences. You can't create a sentence without a verb, and nouns in a sentence are always parameters assigned to something else, unlike verbs which can be freestanding (although they may require nouns or other verbs to be supplied as parameters, as "He" is a parameter of "He rests".)

I don't know a lot about language universals, but another striking feature of verbs in many languages is that they have a variety of syntactic formats, such that many verbs accept multiple formats, but no verb accepts all formats:

  • Intransitive: I eat.
  • Transitive: I eat bread.
  • With a particle: I eat it up.
  • Three arguments: I show the monster a cookie.
  • Three arguments: I put the monster in a cage. (*I put the monster)
  • Verb chain (1): I try to stop.
  • Verb chain (2): I stop eating.
  • Subordinate clause: I hope (that) you're happy

This is in addition to "standard" parameters and complexities that most/all verbs in a language support, such as auxiliary verbs, negatives, and conjugations.

Nouns have several optional parts that can hang off them, but almost any noun (other than verbal nouns, i.e. gerunds) uses the same syntax: the important pen from Earth (note that "pen" could be virtually any noun).

| improve this answer | |
  • 5
    There are a not insignificant number of languages where nouns can form sentences on their own. The language I work on, Tlingit, has sentences consisting entirely of noun phrases despite (or perhaps because of) having an extraordinarily complex verb. Salishan and Wakashan languages can have sentences consisting entirely of nouns, where the nouns even include tense marking. So “sentence forming” is not a valid criterion for noun–verb distinction. – James C. Nov 21 '11 at 19:34
  • 4
    Also, the “syntactic format” you’re bringing up, which is generally referred to as “argument structure”, is a property that is also described for nouns. For example, if a language has an alienability distinction then some nouns have an obligatory argument and some do not. Some nouns even require two arguments, paralleling the intransitive/transitive contrast of verbs. If adpositions are considered to be heads of their own phrases then they too have argument structure, requiring a noun phrase. – James C. Nov 21 '11 at 19:38
  • 3
    @JamesC. As a less sophisticated example, you can consider any language with zero copula (en.wikipedia.org/wiki/Zero_copula); they include languages as unexotic as Latin, Russian or Turkish. – JPP Nov 22 '11 at 10:22
  • @James C, I must ask, when a noun by itself is a sentence, what does the sentence mean? Can any noun be a sentence? – Qwertie Mar 3 '12 at 18:00
  • 1
    I think that generally sentences consisting solely of nouns are equational, with an implicit third person argument. So xóots á-yú ‘brown.bear FOC-DIST’ ‘that’s a bear’. You can analyze the FOC particle as an adjunct of the NP, so the whole sentence is just an NP. There is an implict third person represented by ‘that’ in the English translation, but -yú is not an anaphor and there’s no <i>pro</i> licensed, so there is no syntactic third person argument. You can do this in English sort of, by saying things like ‘bear’ and pointing, but people usually analyze this as elision or the like instead. – James C. Mar 4 '12 at 21:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.